INDEX

Explanations

references to authority, imprisonment, and the consequences of war

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 were

-0.62

Did

-0.61

 weren

-0.58

Did

-0.58

 pinulongan

-0.57

 Were

-0.57

Was

-0.56

 wasn

-0.55

verwijspagina

-0.54

tagHelperRunner

-0.54

POSITIVE LOGITS

can

0.84

is

0.77

čuje

0.75

are

0.71

 cannot

0.71

cannot

0.66

 wordt

0.66

 must

0.63

don

0.62

 darf

0.60

Activations Density 0.581%

references to authority, imprisonment, and the consequences of war

No Comments

No Known Activations